Hibernate Search v5.6.0.Beta1 發布

Hibernate Search的作用是對數據庫中的數據進行檢索的。它是hibernate對著名的全文檢索系統Lucene的一個集成方案，作用在于對數據表中某些內容龐大的字段（如聲明為text的字段）建立全文索引，這樣通過hibernate search就可以對這些字段進行全文檢索后獲得相應的POJO，從而加快了對內容龐大字段進行模糊搜索的速度（sql語句中like匹配）。

Hibernate Search主要有以下功能特點：

1. 功能強大，配置簡單 - 配置只需要修改persistence.xml（JPA），hibernate.cfg.xml（Hibernate）
2. 支持Hibernate，以及EJB3 JPA標準應用
3. 集成全文搜索引擎Lucene - Lucene是Apache項目組下的一個功能強大的全文搜索引擎項目
4. 可以簡單透明索引查詢過的數據
5. 支持復雜檢索 - 支持Wild Card（諸如*, ?等通配符號），多關鍵字，模糊查詢，排序等
6. 支持Clustering
7. 支持直接訪問Lucene API
8. 對Lucene索引，API的高效管理

Hibernate Search運行的環境如下：

1. JDK或JRE 5.0以上
2. Hibernate-Search以及相應的依賴包
3. Hibernate Core 3.2.X
4. Hibernate Annotations 3.3.X

主/從結構的Hibernate Search architecture

 更新日志

Bug修復

• [HSEARCH-2011] - HSEARCH000268 : Field does not exist or not configured for faceting
• [HSEARCH-2046] - Three fields are added to the index for an indexed element collection with two elements
• [HSEARCH-2120] - EdgeCaseFacetTest is unreliable on Elasticsearch
• [HSEARCH-2193] - LuceneBackendQueueTask does not release the Directory lock on update failures
• [HSEARCH-2200] - Typo in log message
• [HSEARCH-2210] - Missing types in NumericFieldUtils#requiresNumericRangeQuery
• [HSEARCH-2222] - The new AnalyzerReference infrastructure is not used in the DSL
• [HSEARCH-2226] - FullTextQuery does not override some returned Query types to allow method chaining
• [HSEARCH-2227] - elasticsearch.ProjectionConstants exposes Lucene constructs
• [HSEARCH-2233] - Elasticsearch - Dates and numerics should never be analyzed
• [HSEARCH-2237] - ScopedAnalyzerReference shouldn't close the underlying analyzer references
• [HSEARCH-2240] - Parallel service lookup might fail to find the service
• [HSEARCH-2251] - In the DSL, don't analyze string encoded dates
• [HSEARCH-2252] - Document that it's not recommended to sort on the id field
• [HSEARCH-2274] - Property "elasticsearch.host" isn't accessed correctly
• [HSEARCH-2275] - Elasticsearch massindexer does not guarantee a proper shutdown
• [HSEARCH-2280] - The JGroups backend module should not depend on the Elasticsearch module

新特性

• [HSEARCH-2064] - Implement filters in Elasticsearch backend
• [HSEARCH-2110] - Upgrade ES documentation before Beta1 release
• [HSEARCH-2199] - Allows the use of CharFilter in the programmatic API of SearchMapping
• [HSEARCH-2234] - Add dependency to "xml.bind" to ES JBoss Modules module.xml

Task

• [HSEARCH-2061] - Review package naming in the new ElasticSearch integration module
• [HSEARCH-2063] - Review the pending TODOs in the new Elasticsearch backend
• [HSEARCH-2067] - Rework ElasticSearchBridgeProvider integration to not affect non-ES indexing
• [HSEARCH-2071] - Support embedded ids with ES backend
• [HSEARCH-2106] - Discuss unique ProjectionConstants API
• [HSEARCH-2138] - Elasticsearch async: do not request reader refresh operations after each write
• [HSEARCH-2156] - Document requiring the bulk deletion Elasticsearch plugin
• [HSEARCH-2164] - Make Elasticsearch configuration properties index-scoped where feasible
• [HSEARCH-2187] - Documentation: Explain how to use the analyzers configured in ES
• [HSEARCH-2198] - Avoid using maven-injection-plugin to retrieve project Version
• [HSEARCH-2205] - Upgrade to Elasticsearch 2.3.1: Java9 compatibility and Lucene aligned to 5.5
• [HSEARCH-2211] - Upgrade Arquillian infrastructure
• [HSEARCH-2217] - Move Travis build to Trusty infrastructure and use Maven wrapper
• [HSEARCH-2229] - Improve error message in ElasticsearchIndexManager#analyzerName
• [HSEARCH-2238] - Upgrade to Byteman 3.0.5
• [HSEARCH-2242] - Convert FacetManager implementation to no longer use Filters
• [HSEARCH-2245] - Handle Calendar and resolution in the Elasticsearch specific BridgeProvider
• [HSEARCH-2246] - Using DSL with Elasticsearch and dates is broken
• [HSEARCH-2264] - Add additional comments on building metadata with/without IndexManager type

Sub-task

• [HSEARCH-2209] - Switch to ClassicSimilarity as default Similarity implementation

改進

• [HSEARCH-1724] - Simplify access to ClassLoaderService
• [HSEARCH-1951] - Improve resulting error message when applying the wrong Sort Type
• [HSEARCH-1970] - Untangle relationship of (DirectoryBased)IndexManager and BackendQueueProvider
• [HSEARCH-2093] - Clarify meaning of constant names on IndexSchemaManagementStrategy
• [HSEARCH-2100] - CPU hotspot in query.dsl.impl.Helper.getDocumentBuilder(QueryBuildingContext)
• [HSEARCH-2125] - Elasticsearch - support for MultiPhraseQuery
• [HSEARCH-2167] - Rename IndexManagementStrategy
• [HSEARCH-2178] - Send bulked index changes to ES when using the mass indexer
• [HSEARCH-2197] - Loading of search results produces useless garbage.
• [HSEARCH-2201] - Fix assorted typos in comments
• [HSEARCH-2202] - Fix SearchIntegrator camel casing throughout the codebase
• [HSEARCH-2203] - Reduce memory consumption of EntityInfoImpl objects.
• [HSEARCH-2212] - Improve Elasticsearch integration documentation
• [HSEARCH-2213] - Cryptic error when cluster is not green during index creation
• [HSEARCH-2214] - Provide an option to wait for Elasticsearch status yellow rather than green
• [HSEARCH-2243] - Allow provided services to implement Startable or Stoppable
• [HSEARCH-2244] - Unify the way TermQuery for numerics are handled
• [HSEARCH-2247] - Represent the kind of underlying technology with an explicit type
• [HSEARCH-2250] - Improve readability and efficiency of MultiWriteDrainableLinkedList
• [HSEARCH-2270] - Upgrade to Apache Lucene 5.5.1.Final

廢棄

• [HSEARCH-2241] - Clarify deprecation of setFilter() method on FullTextQuery

下載

• Source code (zip)
• Source code (tar.gz)